Same here! Actually the Timex Sinclair 1000 was my first computer, but it was followed shortly by the VIC-20 because I wanted something with a real keyboard. Plus I enjoyed the ease of interfacing the VIC to my real world breadboard digital electronics projects.
The sad answer is that yes, this probably is the future of many old computers one day. It's why it is so cool to see people coming up with modern replacement parts that can be retrofitted.
practically any electronics made with capacitors. They just aren't made to last. Even if replacement chips are found, the capacitors are always going to have limited lifespans. I'm not exactly sure why Adrian is so upset about the chips, other than being very low-cost DIY solutions.
@@squirlmy Capacitors are never a problem, you can always replace those and assuming they don't cause damage on the way out, that'll remain the case indefinitely. The issue is critical ICs that have no modern replacements that will eventually fail.
I'm inclined to beleive someone had a number of boards, got as many working as possible by swapping chips around, and these 2 boards are the ones that ended up with all the "suspect" chips, hence the blue spots.
He did say in the Mail Call video (I think) that the person who gave it to him got multiple "for-parts" machines and cobbled together one working unit, then sent Adrian the leftovers. So you're right.
I built a very crude cable descrambler as a kid with a black screen c64 and injecting composite video into the video circuit which provided sync and there was an adjustment to manually sync it. Mostly worked except for the colors of course.
Wow that's pretty neat! It's one thing I never really looked into -- how those scrambled pictures worked and how a descrambler could work. I bet it would be trivial now to build a descrambler -- but I'd like to see a scrambled signal on a scope.
@@adriansdigitalbasement this was known as gated sync suppression. Basically they pushed the sync pulse up into the picture by 6db. There was a subcarrier that would provide the sync so the descrambler could push it back down. By injecting it into the C64 I was restoring the pulses myself. To sync it you would tweak that adj cap you showed on the v20 the c64 had the same thing. It adjusts the “h phase” I believe which does affect the color burst too.
@@adriansdigitalbasement I built a "Universal" descrambler back then from a kit. It used a PLL to lock to the NTSC colorburst signal (which no system ever altered) and used that as timing reference to gate in a new sync pulse in place of the altered one. It also optionally inverted the video so it worked with inverted video methods like Videocipher2 (analog satellite).
First computer I ever touched was a Vic 20 on math class, 6th grade. First computer I owned was a Vic 20. My uncle gave it to me with a tape drive and books and all sorts of goodies. I couldn’t sleep for days just tinkering and typing in “kamakaze Klingon “ line by line in basic: and playing lunar lander forever. Man I miss my Vic.
a 6502 system will usually do 'something' with just the cpu, clock, something to hold the reset vectors (kernal rom or other rom that is either mirrorred up to ffff or actually is at ffff, and preferably stack and zeropage. the vic-20 is a somewhat special case in that the cpu is isolated from the address and databus by some latch logic chips (obviously to let the vic do it's thing). which after pulling everything else out are usually the primary suspects. everything seems to work fine up until setting the border and background color (which on the vic-1 is a combined register holding both) that blue is not the default reset state. funny thing. what makes commodore computers so reliable is that there is absolutely zero error and backchecking. :P 'oh hey imma gonna pull all the i/o chips out and neither the kernal nor the basic will do as much as print an error about it' :P it just initializes stuff and moves on. lol.. it doesn't even properly test the ram :P
I love your videos! Great content almost every single day, im literally in love with retro computers but I can't afford one for myself so I have to live vicariously through your adventures.
I am working on a 6526 replacement. A have a version that can replace an individual chip or can replace both chips on an older 64 where the two chips are next to one another. Will have to wait until the current chip shortage is resolved before I can release to the public. I am hoping that it should cost less than a real 6526 goes for on ebay.
Interesting. I had visited Commodore just as the C64 was beginning development and saw talked with some of the VIC20 and C64 designers. Quite interesting. You mentioned a "Signetics" 6502 but I believe that would be a Synertek 6502 as Synertek was the second source for 6502 chips. I also don't recall Signetics making a 6402. Synertek also developed some of the 6502 peripheral chips such as the 6545 CRTC (similar to the Motorola 6845 CRT Controller chip). At the time I visited, MOS Technology was considering producing the 6545 but I don't believe they ever did. Great video, thank you.
I just picked up a Vic 20 yesterday - along with Radar Ratrace (go figure). No video output (with or without cart) and no sound with cart. Additionally pretty much no chips are socketed. Only 3 sockets in fact: UE8,UD8 and UF8 (the 3 bus transceivers) The Kernal ROM gets very warm, basic ROM gets warm, I think everything else is very cool to the touch. No logic testing equipment at the moment and no vacuum desoldering station (lost in the mail months ago - ugh). Time to follow Ray Carlsen's best advice on Commodore repairs: "if you have a major problem with your VIC20, it's probably best to start looking for another one"
@@jbinary82 There are two ways to produce an FPGA chip, you either analyse the inputs and outputs and manually recreate that, which is similar to a clean room BIOS; or you can look at the actual schematics and recreate what was physically there in the layout.
@@jbinary82 FPGAs on their own being much more advanced and expensive than the machines they are meant to replicate. It's like using an emulator, you get the machine without any of the insight of the containing hardware.
Wow, anyone playing the drinking game is really going to be hammered after this one! It would be really good if you could find a way to test the individual chips on the faulty RAM SIMMs, at least any 4M*4 chips that could be reused on my Let's Make Some Memories SIMMs. I'm wondering if there's a SOJ ZIF socket adapter that could be used with the Retro Chip Tester Pro to test them? I don't think they ever made an adapter for the RAMCHECK that works with the larger capacity chips, but even if they did getting a hold of one would be next to impossible. I have been working on porting my SIMM design to KiCad which will make them fully open-source. But it might be some time since I've been working on other projects, and I want to get a batch made up so I can fully test them before releasing the new version anyway. Then depending if I can find some 16M chips, I might start working on a new board for them.
Good to know not *ALL* of the "Blue Dot" chips were bad. I was about to call it the "Blue Dot of Death!". The issue of chips eventually self-degrading themselves into death is a sad possibility.
I have had at least ten C64's with power-on reset troubles, staying low on pin 40 on CPU. A couple of days ago, same fault on a VIC20. Usally a 7406 TTL logic causes this. U8 in a C64. Sadly, VIC chip is a common fault in VIC20. Kernal ROM is also a common fault, but with a little adapter, it is easy to peplace with a 27C64. I have never seen a dead 6502 or char ROM in a VIC20 before. Usally MOS 85xx chips are more reliable than older MOS 65xx. Thanks for another interesting video :)
Adrian my two Vics (1981 Japan PET and W.German Euro) had two bad faulty vic chips dying 6 months apart after 40 years of proper service , which tells me that the materials probably last around 4 decades. It’s the life cycle of these manufactured chips which will eventually give up and their is nothing we can do about it until it can be replaced. That goes with all vic 20 chips and 64’s. most of the time the power supply outlasts the vast amount of chips which in this case had applied to mine, I had learned this the hard way.
I have a feeling no one ever expected these machines to last beyond the early 90s. Our society has this idea that everything is upgraded in time. I grew up with the Vic 20 and definitely wasn’t interested in using it beyond 1985 or so.
Great video as always! ALL ICs ever made will, one day, stop working, even if they aren't used; guaranteed. It happens because of electron migration, which is faster under usage but can be induced even by cosmic rays. All ICs will die, eventually. MOS chips die a lot faster than the rest because MOS was the worst manufacturer in history when it came to process control (which they basically didn't have) and electron migration mitigation (which they basically didn't do). They sort of worked under the premise that, if the IC worked coming out of the line, then it was good enough; something every fab in history (except for MOS) has known isn't enough.
Bill Herd famously and repeatedly says as much: These parts were never intended nor expected to last 40 years. He's shocked that any of them still work today because zero consideration was ever given to them lasting more than 5-10 years at the most.
So even if you kept working ICs safely stored in a secure airtight container in a vault or whatever for many years without using them they would still break at some point? Is this the same for more modern ICs like processors that we have today? Forgive my ignorance but I just learned about electron migration. I know ICs eventually fail, but didn't realize exactly what ends up killing them.
Love your troubleshooting work. I'm learning. I have this Canadian vic20. Missing a the f7 key. Otherwise good shape. When powered on garbled screen and blind commands work. The two apparently new vic chips, won't do anything. Not even black screen. Could be anything apparently. Lol. I'll I have is a multimeter. Your input if possible would be much appreciated. No. Rush. Happy. Agee, let's keep em going as long as we can
I think in the future we'll have more modern replacements for the chips on those retro computers. I don't think the vintage computers will go away for a long time yet.
Now that there are 6502 to 6510 converters out there such as go4retro's Nu6510 and Monotech's MOS CPU Replacer, hopefully someone will start working on a 6522 to 6526 converter. Given how many C64s are still out there, there's definitely a market for such a device. As for 6502 failures, they're uncommon but not completely unheard of. I've come across one failed 6502 in my travels, along with several failed 6522s. Oddly enough, all of these were made by Rockwell and I found them in BBC Micros.
An 8520 from an Amiga will work in place of a 6526 in a pinch. Not 100% compatible, but very close. Also a dead 1541 is a great place to harvest 6522's.
Another enjoyable video. It seems to me that, unless one has a stash of known good chips, diagnosing and fixing problems on these boards would be rather difficult.
The thing with semiconductors is that they do fail over time. At least these old chips won't fail as quickly as modern chips. But they will sooner or later inevitably stop working. The chip can even be laying on a shelf, packed in anti-static foam, in an argon filled hermetically sealed box, and it will still eventually die of age related failure. One big reason for this is due to diffusion of unwanted dopants entering the silicon. Diffusion is how dopants go into the silicon to start with, usually assisted by having the whole wafer at the time heated to a couple hundred degrees C. A rough rule of thumb is that diffusion gets 2x faster for every 10 degrees C. As an example, a quick 15 minute annealing step at 300 C is going to diffuse these dopants as if one did the same diffusion process at room temperature for 5 thousand years. (This is an excessive oversimplification, but atoms move even if we can't see it with our eyes.) Different materials diffuse at different speeds. Copper for an example is a speed daemon compared to tungsten. Both these materials are however going to kill semiconductors by turning them into resistors as they diffuse into the depletion region. Tungsten is also what makes contact to the transistors themselves, but due to its insanely slow diffusion, it won't reach that depletion region before copper ions in the surrounding packaging has migrated in and done the job. (since most IC packages have some verity of tinned, copper plated steel.) Now, chips have diffusion barriers to stop external contamination from getting into the chip. But diffusion barriers are more like speed restrictions. Its a bumpy gravel road with more potholes than road compared to the smooth flat highway, but it won't stop copper from making it through. It will take them longer, a few decades or perhaps a couple of centuries, but they will get there. (at room temperature, at operating temperatures it goes faster. But old chips like these don't get all that hot in operation.) Modern chips however make a huge sacrifice that these older chips didn't. And that is copper interconnects in the chip. Older chips used aluminium, and aluminium is a weak P dopant and doesn't migrate further than the depletion region. It does slightly change the specifications of the transistor, but outside of precision analog applications the difference is rather meaningless. (also this weak P dopant is why one uses tungsten to contact the transistors, aluminium would introduce an extra diode when connected to N doped silicon, and this would make a lot of circuits not work. Tungsten meanwhile is so slow at diffusing that it doesn't care about becoming a dopant.) Even a lot of modern chips still use aluminium. It has both a reliability advantage, it isn't toxic to the transistors. And it is also cheaper to work with, since it isn't required to be encapsulated in a diffusion barrier. As a metal it is however having ~57% higher resistance so for very power dense applications it isn't always a suitable option. Copper though has the disadvantage that it needs its diffusion barrier, adding cost. And this diffusion barrier can't be too thick, then one ends up with higher overall resistance than if one just used aluminium and then it's pointless to use copper. The diffusion barrier also needs to be thick enough, else copper will meaningfully diffuse through within the product's intended life span. So it isn't atypical that the smallest interconnect layers closest to the transistors are made with aluminium, since the diffusion barrier proportionally takes up too much space to make the overall resistance lower than that of aluminium. While the larger interconnects further up use copper, since the diffusion barrier is proportionally thin enough to make the overall resistance lower than that of aluminium. In practice, a lot of chips don't use copper interconnects. It just isn't advantageous if one has a lower power density, and or aims for a longer product lifespan. In the end of this long comment. All chips will inevitably fail from contamination. More power dense chips made from the late 90's on wards are even more likely to fail in a relatively "short" time span due to the use of copper interconnects. (and also why server CPUs intended for 24/7 operation have lower "max" rated temps for their specifications compared to consumer parts expected to run a lot fewer hours per day.) But I honestly expect there to be proportionally more (or just overall) computers from the 80's alive and running than computers from the late 90's if we look into the future some 30-40 years. (Though, early 2000's systems are likely going to be even more rare as thermal limits weren't really a thing and these chips runs hot enough to need that. So user induced failure is a high possibility.)
MOS really didn't make the most reliable parts, worked in a computer shop in the 90s and we did board repair (well, we swapped in working motherboards so the customer could leave happy and then fixed theirs to go back on the pile). Breadbin Commodore machines made up the vast bulk of repair work. 80% of the time it was a core MOS chip that had just given up the ghost, nothing special on customer side, just it worked right up to the moment it didn't. We didn't even bother testing PSU's back then. Only reason store had a trade in scheme was to source parts for repairs.
You are not kidding about removing cartridges from the vic-20. I still remember clear as day over 30 years later. I had lunar lander and some text based vampire game.
Do you think it would make sense to build a drop-in replacement for the 6502 or 6510 based on the W65C816 (the chip from the Apple II GS)? The idea being to retain full compatibility with existing 8-bit software on Commodores, Ataris, Apples and so forth, while permitting new code to take advantage of the 16 bit processing and 16 MB address space of the 816.
I am curios about something. You went to checking out the chips first instead of looking for bad components on the board like capacitors, resisters, etc.. What made you decide that the problems were mostly likely the chips?
I'm hoping for the whole homemade stuff sector going to "homebrew chips" eventually. Like it did "3D printing" recently; so maybe, eventually, you can just make your own replacement chips for these early ones; the density of the chips might be within a homebrew setup. As long as the actual layout can be found somewhere... so better keep the broken ones around, they might be useful for reverse engineering those 😁
This can basically be done through FPGA progamming. The MiSTer project is specifically: Old arcade electronics replicated by "schematic description languages", and made playable, not just from software, but software on replica hardware schematics on modern hardware.
@@squirlmy Indeed... but to me that feels a bit like cheating 😊 While I might not be the purist that some others are, I do prefer proper hardware based replacements for faulty hardware. Because programming a chip to simulate older hardware makes it feel slightly too much like "emulation" to me already. Where is the boundary? I do see some homebrew chip design projects pop up here and there and hope that those might gain traction eventually. Certainly not "big buck" area, but "special interest" for sure. Kind of the next generation nixie tube thing...?
We were having an extremely rare California tornado. I put on this video and my youngest just sat down and watched the video. Yeah he's a tiny retro nerd. Now that the storm passed he's playing on his powermac all in one.
Hi Adrian, love the content! I just bought some Deoxit myself but every time I spray it, after I stop, it starts dripping down the side of the can/out of the straw (depending on whether the straw is up or down). Have you had that problem before?
I have that with similar bottles. I just make a habit of having a rag or tissue or something to catch it. The valve controls are behind the straw while you’re using it from the straw so there’ll always be a little leftover.
This is a flaw of their new "perfect-straw" as Deoxit call it in their advertising, it is simply awful! You can try to find a regular straw off some other can and see if it fits... They have eventually listened to all the bad feedback on their straw (check the Amazon reviews in Adrians link in this videos description for example!) and now offer the can with the old original straw, part number: D5S-6-LMH Here is a video showing the issue: ua-cam.com/video/rIgFwm1_HUI/v-deo.html
@@perhansson6718 Good to know, thanks. I guess when I run out I'll get that type instead. I wonder how that got past QA, it's not hard to reproduce (literally just use the can as intended)...
@@aaronjamt I suspect it’s deliberate, like how eye drops are metered to be about double what you really need so you buy more sooner. (They _could_ make it so you don’t blink half of a dose out… but they don’t.)
The easiest way to check if the BASIC ROM is causing issues is to put a game cartridge in. When a VIC boots, it tries to run code at #A000 (where a game cartridge sits in memory) and runs that instead of running BASIC if it finds it. Because of that, a machine with just a faulty BASIC ROM will run the game.
Yup, this is exactly what is happening to mine, and boot cartridge runs fine. Boting withought a cartridge, or with a cartridge that dosn't auto run reults in a screen full of garbage. Looks like I do need to swap out that BASIC ROM. I have an eprom programmer but right now I only have a bunch of chips I was using to burn my Vectrex games to. So if I get a compatable chip can I copy the BASIC ROM from my good VIC then burn it to an eprom to work in the BROKEN VIC?
@@GarryGri yes, but it’s not quite that easy. If you could find a 2364 eprom, you can just burn it and drop it in, but those aren’t made anymore and are hard to find. You are much more likely to find 2764s, but then you need some sort of adapter as the pin out isn’t the same. You can either build one (Adrian has shown how to in a previous video), or you can buy adaptor boards on eBay.
@@Doug_in_NC So by the time I buy the 2764 chip and buy/make the adapter is it going to be much cheeper than buying a replacement BASIC ROM chip on ebay for around £20?
@5:20 interesting, The Vic 20 uses a bidirectional bus transceiver 74LS245 for buffering the address lines but forced into one way by grounding the DIR pin. I guess the pin out is straight through tnstead of using a 74LS367 trisatate hex buffer ( and you would need three of them instead of 2)
I managed to acquire what looks like a model 1 Rev b VIC-20. It has 10 ram chips/ gold label and 2 prong adapter. The one issue I'm concerned about is heat. It has a large black heatsink that gets hot to the touch. Is this normal? The main cap looks a bit crusty which needs changing out. thanx
Great video AB! I really miss myy VIC-20 and all the neat memory/BASIC/Video configuration address spaces. The video makes me curious as to what was the deal with the blue stickers on all those chips...could it be they were questionable to begin with? C= did some odd stuff around the time JT left...Thanks again!
@@dalriada842 Deoxit famously offers to send almost all tech UA-camrs (even pretty small ones) free cans so long as they always say Deoxit and never the generic “contact cleaner”. VWestlife and Techmoan have talked about turning them down but I generally assume most folks who regularly use it and say the name every time have probably accepted the offer (which likely includes not talking about it in the terms).
@@dalriada842 they usually email UA-camrs once they’ve already been seen using Deoxit - like “hey, want a free lifetime supply?” which also pre-screens for people who already like it. But at least in Adrian’s case I don’t think he’s hamming it up even if he is getting freebies - it’s the comments and mail-call letters who’ve come up with phrases like “Deoxit that socket”, so there’s a symbiotic relationship between the payment-in-kind (if indeed he is getting the freebies) and audience reaction. Which feels different to me than top-down trying to “make it a thing”. Like, he mentions spraying Deoxit a couple years ago but called less explicit attention to it (and Americans tend to use brand names more often anyway so there’s no easy way to infer when they may have emailed or not).
The other 6500 series ICs were made by Synertek, not Signetics... I only mention it because Synertek supplied tons of chips to Atari, Apple, etc. but they never get any love :( ...also their chips seem to be much more reliable than MOS Technology :)
Hello Adrian, I was wondering what resource you use to get switches and keycaps for the Vic20. I've been a fan and subscriber to your channel for some time now. I'm a Tandy CoCo guy but always wanted a Commodore Computer for my collection. Thank you for any info you can provide me. ebay suppliers just give you random keycaps and switches.
For those dead VIC chips if I were to make a guess of why these die would be heat. They all get hot inside of a closed metal box. With my VIC-20 systems the first thing I always do is open them, take off that VIC video chip box cover and add a heatsink....if it is working when I get the machine obviously. Then I heatsink the rest. Personally however I have never actually encountered a dead VIC-20. I perform the preventative maintenance, but again they are never dead. They always however need pot adjustment. Even so, the chips are old and will all fail eventually no matter what we do. It would be nice to see VIC video replacements though like we see on the C64 with things like the Kawari.
I wonder how many of those memory sticks in the dead-box could be revived by swapping chips/reworking them? (providing you could trace the faults to specific chips) Also, what memory-testing program are you using to ferret out bad sticks?
I bought my cousin's old Vic 20 years ago (1990) He couldn't remember if it worked. We connected it, & it kept saying "boot error" or "reboot error" & neither of us knew how to deal with it I just packed it away along with a C64 with Monitor given as payment for helping clean out a closed down very dirty Flea Market. They're available. I also got games and storage device, joystick, all very old, in the original dusty scratched box.
I've not worked on a VIC-20 before... but I've got two failed MOS 6502 CPUs that came out of my original 1541 and also an SX-64 Floppy Board that I'm working on. Both presented the same symptoms, activity LED lit all the time and the floppy drive motor spinning. Both CPUs also got VERY hot, too hot to touch.
Future troubleshooting methodology: after checking power supply and PCB for proper voltages, connect sound, video, keyboard, and insert game cart to see what is actually working. The presence of keyboard control and sound are great clues as to what is wrong and what isn't.
Great :) I guess all chips will eventually die. But in my experience, its usually chips that get warm die the fastest. Those are the first to give up. Maybe its heat stresses over time, whatever, but they seem to go that way. Never seen a 6502 go bad yet, thats a first to me!.
A big hint that the VIC was bad was the cyan screen and then yellow for "Raid on Fort Knox" and then blue for "Radar Ratrace". The VIC started drawing the border and never stopped.
It is important to know that the newer VIC20 _can_ use the standard C64 power supply, but its original brick _must not_ be used with the C64. It will work for a while but the regulator for the 5V rail is not designed for the load, and will eventually fail destructively.
Sucks when the VIC-II dies, it goes double for the VIC I, because those are WAY harder to find (The NTSC version at least). I spent $99 a pop for the 2 6560R2's I bought
A 6502, on coming out of reset, goes to the top two addresses in memory ($fffe and $ffff) to get the startup vector, the address to jump to to begin system initialization. One a Commodore machine, you can get surprisingly far without working RAM, a PET can make it's power on bleeps without needing working RAM. This is rather unusual given how dependent on zero page and stack operations most 6502 code is.
I wonder if you could do some kind of analysis to see if those blue dot parts would have all been damaged by the same event, like a bad PSU outputting too much voltage or something.
How much do I have to bump up my patreon to send in my black screen Vic-20 early 300 series assy that i recapped and replaced the volt regulator- but don't have the tools/skills to check every IC?? lol😁
Been trying to repair my childhood Vic-20 for a while now. Aside from a bad reset circuit capacitor, which was holding _RESET low for 30 seconds(!) before boot, bad 6560-VIC. Got a couple replacements from kind people on Reddit, but they both had problems as well. First one had a bad clock divider, so no clock to the CPU. Second one can't seem to consistently drive the address bus to get character data and is a thermal fault. Cold Spray helps, but not for long. Makes me worried that all VIC-I chips' days are numbered like the PLA on the C64 or the CPU on the C16/Plus4.
40:00 people have been working on rudimentary IC fabrication at home but perhaps before that becomes feasible to produce retro replacement parts at scale we might get companies starting to provide cheap custom FlexICs (ICs on flex-cable type substrate, they already made an implementation of 6502 on it). all we'd need is for someone to reverse-engineer any missing chips
well flexIC might not be suitable for making replacement parts for silicon circuits in foreseeable future because it can't switch as fast (the flex6502 runs at 71khz) but maybe some day
28:12 I remember reading that they made changes to the Commodore 64's KERNAL over the years, and one change was that in some versions the entire color ROM was initialized to light-blue (so if you were to POKE a character into an empty space, it would be visible), but in others it defaulted to the background color and only made the visible characters light blue. I guess you've only ever run this test on machines that used the first style.
I have a merch suggestion - it's making a deoxit that socket shirt (I have a light blue "It Freakin Works" shirt - Spring sucks with their printing process as the ink rubs off easily even with my chewed down not sharp for anything fingernails but Spring is shipping me a new one)
The S logo seems to be the Synertek logo, not Signetics. I don't think Signetics actually did 6502/6522 stuff. The scary thing is the 6522 is still made as the WDC 65C22 and the N variant is directly compatible! (edit) It seems WDC also make a 65C02 which is compatible
The MiSTer project addresses this, as FPGAs are literally programmed using Hardware Description Languages. And someday (perhaps when patents and trademarks expire), the chips will be both be able to be re-created from VHDL or perhaps some other hardware description language, or likely the more practical solution; continue to run as designed in "virtual" FPGA solutions. One additional issue for Commodore computers are the keyboards, but even now, TheC64 can be hacked to work as a USB keyboard.
the 6502 always is a problem on VIC20 (well if you have experience, that is) and not so much on 1541. ROM also is known to go bad on VIC20, 6560/61 also seems to go bad sometimes. Just usualy VIC20 stuff.
It could be that the guy who sent you these, populated the sockets with questionable parts. He did say these were left overs after he repaired his. You could place appropriate TVS on the supply lines, to make sure you don't get over-voltage spikes.
over voltage spikes would be a power supply problem... best thing to do there is use modern power supplies rather than the old junk that they shipped with
MOS had questionable quality in general, from their wafer repair technology that turned defective chips marginal. And the mentioned 6526 are especially flimsy by nature.
@@Breakfast_of_Champions yes! I had said to my screen “that’s from their secret sauce that improved yields” :)the older chips had to be perfect to pass inspection so it makes sense they’re hardier.
The fact that the song is out of tune in Rat Race makes it even better. You said it annoyed you as a kid but I still kinda laugh at the notion that your parents probably hated it more and regretted buying it. I can imagine one or both sitting down to relax when they hear the first three notes in the distance and then cringing because they know the out of tune rendition of Three Blind Mice will repeat for what feels like forever. Did you use it as a tool/weapon to get more games? Did that cartridge mysteriously disappear at times or suddenly not work for some reason? Or perhaps Rat Race hastened the purchase of a C64. Yes, I can imagine it all.
Vic 20 mini-repairathon! Now we are talking!!!!!
Hearing "Deoxit that socket" made the old "don't copy that floppy" thing flash into my mind.
Since it was my first computer, the VIC-20 has a special place in my heart. It's always good to see systems that are able to be saved.
Mine was the c64 but always wanted a vic-20 myself. Those seemed really underrated.
Same here! Actually the Timex Sinclair 1000 was my first computer, but it was followed shortly by the VIC-20 because I wanted something with a real keyboard. Plus I enjoyed the ease of interfacing the VIC to my real world breadboard digital electronics projects.
As a huge Vic 20 fan, I appreciate that you were able to save a couple of these machines. It was a good watch. Thank you.
The sad answer is that yes, this probably is the future of many old computers one day. It's why it is so cool to see people coming up with modern replacement parts that can be retrofitted.
The VIC is relatively fortunate in that aspect as there are still 6502 and 6522 variants made. We just need a modern VIC replacement of the big chips.
practically any electronics made with capacitors. They just aren't made to last. Even if replacement chips are found, the capacitors are always going to have limited lifespans. I'm not exactly sure why Adrian is so upset about the chips, other than being very low-cost DIY solutions.
@@squirlmy Capacitors are never a problem, you can always replace those and assuming they don't cause damage on the way out, that'll remain the case indefinitely. The issue is critical ICs that have no modern replacements that will eventually fail.
I'm inclined to beleive someone had a number of boards, got as many working as possible by swapping chips around, and these 2 boards are the ones that ended up with all the "suspect" chips, hence the blue spots.
It's like the one egg carton at the supermarket that ends up with all the broken eggs.
That sounds more feasible, because that was just too many failed ICs on one MB.
He did say in the Mail Call video (I think) that the person who gave it to him got multiple "for-parts" machines and cobbled together one working unit, then sent Adrian the leftovers. So you're right.
@@mal2ksc remind me not to shop at the supermarket you frequent!
@@HardDriveGuruOfficial although your interpretation is more sinister!!! A good reason to avoid "untested" and "for parts" purchases.
Starting to look like you might have a use for a ZIF-20 as well as a ZIF-64!
Thank you for the cool video. I just got my first VIC a few weeks ago with 8 games, 6 cassettes, and a C2N that I cannot wait to get booted up.
I do like this channel. You explain everything and you make mistakes that everyone makes and most of all you are clear and concise.
I built a very crude cable descrambler as a kid with a black screen c64 and injecting composite video into the video circuit which provided sync and there was an adjustment to manually sync it. Mostly worked except for the colors of course.
Wow that's pretty neat! It's one thing I never really looked into -- how those scrambled pictures worked and how a descrambler could work. I bet it would be trivial now to build a descrambler -- but I'd like to see a scrambled signal on a scope.
@@adriansdigitalbasement this was known as gated sync suppression. Basically they pushed the sync pulse up into the picture by 6db. There was a subcarrier that would provide the sync so the descrambler could push it back down. By injecting it into the C64 I was restoring the pulses myself. To sync it you would tweak that adj cap you showed on the v20 the c64 had the same thing. It adjusts the “h phase” I believe which does affect the color burst too.
@@adriansdigitalbasement I built a "Universal" descrambler back then from a kit. It used a PLL to lock to the NTSC colorburst signal (which no system ever altered) and used that as timing reference to gate in a new sync pulse in place of the altered one. It also optionally inverted the video so it worked with inverted video methods like Videocipher2 (analog satellite).
First computer I ever touched was a Vic 20 on math class, 6th grade. First computer I owned was a Vic 20. My uncle gave it to me with a tape drive and books and all sorts of goodies. I couldn’t sleep for days just tinkering and typing in “kamakaze Klingon “ line by line in basic: and playing lunar lander forever. Man I miss my Vic.
I'm so touched by Adrian telling us about his bad memories 😢😢😢
a 6502 system will usually do 'something' with just the cpu, clock, something to hold the reset vectors (kernal rom or other rom that is either mirrorred up to ffff or actually is at ffff, and preferably stack and zeropage. the vic-20 is a somewhat special case in that the cpu is isolated from the address and databus by some latch logic chips (obviously to let the vic do it's thing). which after pulling everything else out are usually the primary suspects. everything seems to work fine up until setting the border and background color (which on the vic-1 is a combined register holding both) that blue is not the default reset state. funny thing. what makes commodore computers so reliable is that there is absolutely zero error and backchecking. :P 'oh hey imma gonna pull all the i/o chips out and neither the kernal nor the basic will do as much as print an error about it' :P it just initializes stuff and moves on. lol.. it doesn't even properly test the ram :P
I love your videos! Great content almost every single day, im literally in love with retro computers but I can't afford one for myself so I have to live vicariously through your adventures.
The VICS are working. A repair is a repair, regardless. Thumbs up.
Thanks!
Thanks for the support!
VIC20 were very durable and reliable. I used one as an auto-dialer to test 3 pager systems, ran for over 10 years.
I am working on a 6526 replacement. A have a version that can replace an individual chip or can replace both chips on an older 64 where the two chips are next to one another. Will have to wait until the current chip shortage is resolved before I can release to the public. I am hoping that it should cost less than a real 6526 goes for on ebay.
Interesting. I had visited Commodore just as the C64 was beginning development and saw talked with some of the VIC20 and C64 designers. Quite interesting. You mentioned a "Signetics" 6502 but I believe that would be a Synertek 6502 as Synertek was the second source for 6502 chips. I also don't recall Signetics making a 6402. Synertek also developed some of the 6502 peripheral chips such as the 6545 CRTC (similar to the Motorola 6845 CRT Controller chip). At the time I visited, MOS Technology was considering producing the 6545 but I don't believe they ever did. Great video, thank you.
I just picked up a Vic 20 yesterday - along with Radar Ratrace (go figure). No video output (with or without cart) and no sound with cart. Additionally pretty much no chips are socketed. Only 3 sockets in fact: UE8,UD8 and UF8 (the 3 bus transceivers) The Kernal ROM gets very warm, basic ROM gets warm, I think everything else is very cool to the touch. No logic testing equipment at the moment and no vacuum desoldering station (lost in the mail months ago - ugh). Time to follow Ray Carlsen's best advice on Commodore repairs: "if you have a major problem with your VIC20, it's probably best to start looking for another one"
Let's hope some day the designs of the chips will be found and replacements can be made.
There are FPGA reproductions of vic20s, I guess they are already reproduced somehow.
@@jbinary82 There are two ways to produce an FPGA chip, you either analyse the inputs and outputs and manually recreate that, which is similar to a clean room BIOS; or you can look at the actual schematics and recreate what was physically there in the layout.
@@jbinary82 FPGAs on their own being much more advanced and expensive than the machines they are meant to replicate.
It's like using an emulator, you get the machine without any of the insight of the containing hardware.
Wow, anyone playing the drinking game is really going to be hammered after this one!
It would be really good if you could find a way to test the individual chips on the faulty RAM SIMMs, at least any 4M*4 chips that could be reused on my Let's Make Some Memories SIMMs. I'm wondering if there's a SOJ ZIF socket adapter that could be used with the Retro Chip Tester Pro to test them? I don't think they ever made an adapter for the RAMCHECK that works with the larger capacity chips, but even if they did getting a hold of one would be next to impossible.
I have been working on porting my SIMM design to KiCad which will make them fully open-source. But it might be some time since I've been working on other projects, and I want to get a batch made up so I can fully test them before releasing the new version anyway. Then depending if I can find some 16M chips, I might start working on a new board for them.
Great show Adrian I have a non working VIC20 so I can now make a move on getting it to go properly .
Good to know not *ALL* of the "Blue Dot" chips were bad. I was about to call it the "Blue Dot of Death!". The issue of chips eventually self-degrading themselves into death is a sad possibility.
or blue dot Easter eggs 😀
I have had at least ten C64's with power-on reset troubles, staying low on pin 40 on CPU. A couple of days ago, same fault on a VIC20. Usally a 7406 TTL logic causes this. U8 in a C64.
Sadly, VIC chip is a common fault in VIC20. Kernal ROM is also a common fault, but with a little adapter, it is easy to peplace with a 27C64.
I have never seen a dead 6502 or char ROM in a VIC20 before.
Usally MOS 85xx chips are more reliable than older MOS 65xx.
Thanks for another interesting video :)
Had one SX64 with failed reset signal. Culprit was 7416. So yes, for some reason even this small circuit is failing sometimes.
Local man goes mad with power, starts swapping integrated circuits with wild abandon, coats things in Deoxit, d̶o̶e̶s̶n̶'̶t̶ elaborates.
Adrian my two Vics (1981 Japan PET and W.German Euro) had two bad faulty vic chips dying 6 months apart after 40 years of proper service , which tells me that the materials probably last around 4 decades. It’s the life cycle of these manufactured chips which will eventually give up and their is nothing we can do about it until it can be replaced. That goes with all vic 20 chips and 64’s. most of the time the power supply outlasts the vast amount of chips which in this case had applied to mine, I had learned this the hard way.
I have a feeling no one ever expected these machines to last beyond the early 90s. Our society has this idea that everything is upgraded in time. I grew up with the Vic 20 and definitely wasn’t interested in using it beyond 1985 or so.
Great video as always!
ALL ICs ever made will, one day, stop working, even if they aren't used; guaranteed. It happens because of electron migration, which is faster under usage but can be induced even by cosmic rays. All ICs will die, eventually.
MOS chips die a lot faster than the rest because MOS was the worst manufacturer in history when it came to process control (which they basically didn't have) and electron migration mitigation (which they basically didn't do). They sort of worked under the premise that, if the IC worked coming out of the line, then it was good enough; something every fab in history (except for MOS) has known isn't enough.
Bill Herd famously and repeatedly says as much: These parts were never intended nor expected to last 40 years. He's shocked that any of them still work today because zero consideration was ever given to them lasting more than 5-10 years at the most.
So even if you kept working ICs safely stored in a secure airtight container in a vault or whatever for many years without using them they would still break at some point? Is this the same for more modern ICs like processors that we have today?
Forgive my ignorance but I just learned about electron migration. I know ICs eventually fail, but didn't realize exactly what ends up killing them.
Love your troubleshooting work. I'm learning. I have this Canadian vic20. Missing a the f7 key. Otherwise good shape. When powered on garbled screen and blind commands work. The two apparently new vic chips, won't do anything. Not even black screen. Could be anything apparently. Lol. I'll I have is a multimeter. Your input if possible would be much appreciated. No. Rush. Happy. Agee, let's keep em going as long as we can
I think in the future we'll have more modern replacements for the chips on those retro computers. I don't think the vintage computers will go away for a long time yet.
Now that there are 6502 to 6510 converters out there such as go4retro's Nu6510 and Monotech's MOS CPU Replacer, hopefully someone will start working on a 6522 to 6526 converter. Given how many C64s are still out there, there's definitely a market for such a device.
As for 6502 failures, they're uncommon but not completely unheard of. I've come across one failed 6502 in my travels, along with several failed 6522s. Oddly enough, all of these were made by Rockwell and I found them in BBC Micros.
An 8520 from an Amiga will work in place of a 6526 in a pinch. Not 100% compatible, but very close. Also a dead 1541 is a great place to harvest 6522's.
For easier chip check-off visibility, try the metallic color Sharpies if you haven't already.
yes, the Sharpies that are paint work great on glass and plastic, since I "discovered" them I use them regularly.
He had one mail video, where he was given an white paint Edding specifically for marking the ICs.. I wonder what he did to this.
Another enjoyable video. It seems to me that, unless one has a stash of known good chips, diagnosing and fixing problems on these boards would be rather difficult.
4:52 I've had a bad reset signal, but not in a computer, it was in a DAT audio recorder. The timing capacitor had failed.
The thing with semiconductors is that they do fail over time. At least these old chips won't fail as quickly as modern chips. But they will sooner or later inevitably stop working. The chip can even be laying on a shelf, packed in anti-static foam, in an argon filled hermetically sealed box, and it will still eventually die of age related failure.
One big reason for this is due to diffusion of unwanted dopants entering the silicon. Diffusion is how dopants go into the silicon to start with, usually assisted by having the whole wafer at the time heated to a couple hundred degrees C. A rough rule of thumb is that diffusion gets 2x faster for every 10 degrees C. As an example, a quick 15 minute annealing step at 300 C is going to diffuse these dopants as if one did the same diffusion process at room temperature for 5 thousand years. (This is an excessive oversimplification, but atoms move even if we can't see it with our eyes.)
Different materials diffuse at different speeds. Copper for an example is a speed daemon compared to tungsten. Both these materials are however going to kill semiconductors by turning them into resistors as they diffuse into the depletion region. Tungsten is also what makes contact to the transistors themselves, but due to its insanely slow diffusion, it won't reach that depletion region before copper ions in the surrounding packaging has migrated in and done the job. (since most IC packages have some verity of tinned, copper plated steel.)
Now, chips have diffusion barriers to stop external contamination from getting into the chip. But diffusion barriers are more like speed restrictions. Its a bumpy gravel road with more potholes than road compared to the smooth flat highway, but it won't stop copper from making it through. It will take them longer, a few decades or perhaps a couple of centuries, but they will get there. (at room temperature, at operating temperatures it goes faster. But old chips like these don't get all that hot in operation.)
Modern chips however make a huge sacrifice that these older chips didn't. And that is copper interconnects in the chip. Older chips used aluminium, and aluminium is a weak P dopant and doesn't migrate further than the depletion region. It does slightly change the specifications of the transistor, but outside of precision analog applications the difference is rather meaningless. (also this weak P dopant is why one uses tungsten to contact the transistors, aluminium would introduce an extra diode when connected to N doped silicon, and this would make a lot of circuits not work. Tungsten meanwhile is so slow at diffusing that it doesn't care about becoming a dopant.)
Even a lot of modern chips still use aluminium. It has both a reliability advantage, it isn't toxic to the transistors. And it is also cheaper to work with, since it isn't required to be encapsulated in a diffusion barrier. As a metal it is however having ~57% higher resistance so for very power dense applications it isn't always a suitable option.
Copper though has the disadvantage that it needs its diffusion barrier, adding cost. And this diffusion barrier can't be too thick, then one ends up with higher overall resistance than if one just used aluminium and then it's pointless to use copper. The diffusion barrier also needs to be thick enough, else copper will meaningfully diffuse through within the product's intended life span. So it isn't atypical that the smallest interconnect layers closest to the transistors are made with aluminium, since the diffusion barrier proportionally takes up too much space to make the overall resistance lower than that of aluminium. While the larger interconnects further up use copper, since the diffusion barrier is proportionally thin enough to make the overall resistance lower than that of aluminium.
In practice, a lot of chips don't use copper interconnects. It just isn't advantageous if one has a lower power density, and or aims for a longer product lifespan.
In the end of this long comment.
All chips will inevitably fail from contamination.
More power dense chips made from the late 90's on wards are even more likely to fail in a relatively "short" time span due to the use of copper interconnects. (and also why server CPUs intended for 24/7 operation have lower "max" rated temps for their specifications compared to consumer parts expected to run a lot fewer hours per day.)
But I honestly expect there to be proportionally more (or just overall) computers from the 80's alive and running than computers from the late 90's if we look into the future some 30-40 years. (Though, early 2000's systems are likely going to be even more rare as thermal limits weren't really a thing and these chips runs hot enough to need that. So user induced failure is a high possibility.)
Adrian’s new video… sweet!!!
MOS really didn't make the most reliable parts, worked in a computer shop in the 90s and we did board repair (well, we swapped in working motherboards so the customer could leave happy and then fixed theirs to go back on the pile). Breadbin Commodore machines made up the vast bulk of repair work. 80% of the time it was a core MOS chip that had just given up the ghost, nothing special on customer side, just it worked right up to the moment it didn't. We didn't even bother testing PSU's back then. Only reason store had a trade in scheme was to source parts for repairs.
We need further ado! ^-^
I always enjoy watching these. Keep up the good work!
You are not kidding about removing cartridges from the vic-20. I still remember clear as day over 30 years later. I had lunar lander and some text based vampire game.
Uh, Dracula (I'm not pulling your leg)?
Do you think it would make sense to build a drop-in replacement for the 6502 or 6510 based on the W65C816 (the chip from the Apple II GS)? The idea being to retain full compatibility with existing 8-bit software on Commodores, Ataris, Apples and so forth, while permitting new code to take advantage of the 16 bit processing and 16 MB address space of the 816.
I am curios about something. You went to checking out the chips first instead of looking for bad components on the board like capacitors, resisters, etc.. What made you decide that the problems were mostly likely the chips?
I'm hoping for the whole homemade stuff sector going to "homebrew chips" eventually. Like it did "3D printing" recently; so maybe, eventually, you can just make your own replacement chips for these early ones; the density of the chips might be within a homebrew setup. As long as the actual layout can be found somewhere... so better keep the broken ones around, they might be useful for reverse engineering those 😁
This can basically be done through FPGA progamming. The MiSTer project is specifically: Old arcade electronics replicated by "schematic description languages", and made playable, not just from software, but software on replica hardware schematics on modern hardware.
@@squirlmy Indeed... but to me that feels a bit like cheating 😊 While I might not be the purist that some others are, I do prefer proper hardware based replacements for faulty hardware. Because programming a chip to simulate older hardware makes it feel slightly too much like "emulation" to me already. Where is the boundary? I do see some homebrew chip design projects pop up here and there and hope that those might gain traction eventually. Certainly not "big buck" area, but "special interest" for sure. Kind of the next generation nixie tube thing...?
I realize that's a ram on your dead parts box, but I'd like to think of it as the Icon of SIMM.
We were having an extremely rare California tornado. I put on this video and my youngest just sat down and watched the video. Yeah he's a tiny retro nerd. Now that the storm passed he's playing on his powermac all in one.
Hi Adrian, love the content! I just bought some Deoxit myself but every time I spray it, after I stop, it starts dripping down the side of the can/out of the straw (depending on whether the straw is up or down). Have you had that problem before?
I have that with similar bottles. I just make a habit of having a rag or tissue or something to catch it. The valve controls are behind the straw while you’re using it from the straw so there’ll always be a little leftover.
This is a flaw of their new "perfect-straw" as Deoxit call it in their advertising, it is simply awful! You can try to find a regular straw off some other can and see if it fits... They have eventually listened to all the bad feedback on their straw (check the Amazon reviews in Adrians link in this videos description for example!) and now offer the can with the old original straw, part number: D5S-6-LMH
Here is a video showing the issue: ua-cam.com/video/rIgFwm1_HUI/v-deo.html
@@kaitlyn__L Yeah, that's what I've been doing. I was hoping there's a better solution or it's user error.
@@perhansson6718 Good to know, thanks. I guess when I run out I'll get that type instead. I wonder how that got past QA, it's not hard to reproduce (literally just use the can as intended)...
@@aaronjamt I suspect it’s deliberate, like how eye drops are metered to be about double what you really need so you buy more sooner. (They _could_ make it so you don’t blink half of a dose out… but they don’t.)
Awesome! I have one I need to button up and play with soon!
The easiest way to check if the BASIC ROM is causing issues is to put a game cartridge in. When a VIC boots, it tries to run code at #A000 (where a game cartridge sits in memory) and runs that instead of running BASIC if it finds it. Because of that, a machine with just a faulty BASIC ROM will run the game.
You might still want to pull the BASIC ROM, in case it has bad chip select logic and is driving the bus (unlikely, but...).
Yup, this is exactly what is happening to mine, and boot cartridge runs fine. Boting withought a cartridge, or with a cartridge that dosn't auto run reults in a screen full of garbage.
Looks like I do need to swap out that BASIC ROM.
I have an eprom programmer but right now I only have a bunch of chips I was using to burn my Vectrex games to. So if I get a compatable chip can I copy the BASIC ROM from my good VIC then burn it to an eprom to work in the BROKEN VIC?
@@GarryGri yes, but it’s not quite that easy. If you could find a 2364 eprom, you can just burn it and drop it in, but those aren’t made anymore and are hard to find. You are much more likely to find 2764s, but then you need some sort of adapter as the pin out isn’t the same. You can either build one (Adrian has shown how to in a previous video), or you can buy adaptor boards on eBay.
@@Doug_in_NC So by the time I buy the 2764 chip and buy/make the adapter is it going to be much cheeper than buying a replacement BASIC ROM chip on ebay for around £20?
I just saw one on ebay for a tenner NOS, so I'll buy it and try it!
@5:20 interesting, The Vic 20 uses a bidirectional bus transceiver 74LS245 for buffering the address lines but forced into one way by grounding the DIR pin. I guess the pin out is straight through tnstead of using a 74LS367 trisatate hex buffer ( and you would need three of them instead of 2)
You can get some brand new replacements like 6522, and the SRAM, and the 3 color RAMs, and the 245s (using 645s). All new.
I managed to acquire what looks like a model 1 Rev b VIC-20. It has 10 ram chips/ gold label and 2 prong adapter. The one issue I'm concerned about is heat. It has a large black heatsink that gets hot to the touch. Is this normal? The main cap looks a bit crusty which needs changing out. thanx
Radar Rat Race was the first game I had on my VIC20. It brings back so much nostalgia!
Sometimes oxidation just makes the surface rough, that's why it makes chips hard to insert.
Great video AB! I really miss myy VIC-20 and all the neat memory/BASIC/Video configuration address spaces. The video makes me curious as to what was the deal with the blue stickers on all those chips...could it be they were questionable to begin with? C= did some odd stuff around the time JT left...Thanks again!
I made "Deoxit!" into a drinking game and I'm now writing this from an ICU.
They should pay him for advertising their product. I bought a can because he kept raving about it.
@@dalriada842 Deoxit famously offers to send almost all tech UA-camrs (even pretty small ones) free cans so long as they always say Deoxit and never the generic “contact cleaner”. VWestlife and Techmoan have talked about turning them down but I generally assume most folks who regularly use it and say the name every time have probably accepted the offer (which likely includes not talking about it in the terms).
@@kaitlyn__L Interesting! I'd be disappointed if that was the true motive.
@@dalriada842 they usually email UA-camrs once they’ve already been seen using Deoxit - like “hey, want a free lifetime supply?” which also pre-screens for people who already like it.
But at least in Adrian’s case I don’t think he’s hamming it up even if he is getting freebies - it’s the comments and mail-call letters who’ve come up with phrases like “Deoxit that socket”, so there’s a symbiotic relationship between the payment-in-kind (if indeed he is getting the freebies) and audience reaction. Which feels different to me than top-down trying to “make it a thing”.
Like, he mentions spraying Deoxit a couple years ago but called less explicit attention to it (and Americans tend to use brand names more often anyway so there’s no easy way to infer when they may have emailed or not).
The other 6500 series ICs were made by Synertek, not Signetics... I only mention it because Synertek supplied tons of chips to Atari, Apple, etc. but they never get any love :( ...also their chips seem to be much more reliable than MOS Technology :)
Hello Adrian, I was wondering what resource you use to get switches and keycaps for the Vic20. I've been a fan and subscriber to your channel for some time now. I'm a Tandy CoCo guy but always wanted a Commodore Computer for my collection. Thank you for any info you can provide me. ebay suppliers just give you random keycaps and switches.
Did you try Deoxiting those 5 chips?
For those dead VIC chips if I were to make a guess of why these die would be heat. They all get hot inside of a closed metal box. With my VIC-20 systems the first thing I always do is open them, take off that VIC video chip box cover and add a heatsink....if it is working when I get the machine obviously. Then I heatsink the rest.
Personally however I have never actually encountered a dead VIC-20. I perform the preventative maintenance, but again they are never dead. They always however need pot adjustment.
Even so, the chips are old and will all fail eventually no matter what we do. It would be nice to see VIC video replacements though like we see on the C64 with things like the Kawari.
I wonder how many of those memory sticks in the dead-box could be revived by swapping chips/reworking them?
(providing you could trace the faults to specific chips)
Also, what memory-testing program are you using to ferret out bad sticks?
I bought my cousin's old Vic 20 years ago (1990)
He couldn't remember if it worked.
We connected it, & it kept saying "boot error" or "reboot error" & neither of us knew how to deal with it
I just packed it away along with a C64 with Monitor given as payment for helping clean out a closed down very dirty Flea Market.
They're available. I also got games and storage device, joystick, all very old, in the original dusty scratched box.
I've not worked on a VIC-20 before... but I've got two failed MOS 6502 CPUs that came out of my original 1541 and also an SX-64 Floppy Board that I'm working on. Both presented the same symptoms, activity LED lit all the time and the floppy drive motor spinning. Both CPUs also got VERY hot, too hot to touch.
22:13 “pop out this cartridge” lol pop out a Vic 20 cart
Will you do an Amiga (500) repairaton once? would be awesome!
Hi
Do you have a linkt to the video in the intro where you wash the board?
Future troubleshooting methodology: after checking power supply and PCB for proper voltages, connect sound, video, keyboard, and insert game cart to see what is actually working. The presence of keyboard control and sound are great clues as to what is wrong and what isn't.
Great :) I guess all chips will eventually die. But in my experience, its usually chips that get warm die the fastest. Those are the first to give up. Maybe its heat stresses over time, whatever, but they seem to go that way. Never seen a 6502 go bad yet, thats a first to me!.
A big hint that the VIC was bad was the cyan screen and then yellow for "Raid on Fort Knox" and then blue for "Radar Ratrace". The VIC started drawing the border and never stopped.
I miss my old VIC20, wish I still had it. Also it's sad that Adrian has a box full of bad memories...😜
I wonder if certain climates are more conducive to chip failures.
It is important to know that the newer VIC20 _can_ use the standard C64 power supply, but its original brick _must not_ be used with the C64. It will work for a while but the regulator for the 5V rail is not designed for the load, and will eventually fail destructively.
No VIC-20 diagnostic program!? Blasphemy!
Jk, awesome content as always, Adrian, Rammy, and Digital Basement. 👏🏻👏🏻👏🏻
Sucks when the VIC-II dies, it goes double for the VIC I, because those are WAY harder to find (The NTSC version at least). I spent $99 a pop for the 2 6560R2's I bought
What’s this RAM tester you spoke of at the end?/
A 6502, on coming out of reset, goes to the top two addresses in memory ($fffe and $ffff) to get the startup vector, the address to jump to to begin system initialization. One a Commodore machine, you can get surprisingly far without working RAM, a PET can make it's power on bleeps without needing working RAM. This is rather unusual given how dependent on zero page and stack operations most 6502 code is.
I wonder if you could do some kind of analysis to see if those blue dot parts would have all been damaged by the same event, like a bad PSU outputting too much voltage or something.
Perhaps they was plugged in during a storm. Lightning can do some bad stuff
How much do I have to bump up my patreon to send in my black screen Vic-20 early 300 series assy that i recapped and replaced the volt regulator- but don't have the tools/skills to check every IC?? lol😁
Been trying to repair my childhood Vic-20 for a while now. Aside from a bad reset circuit capacitor, which was holding _RESET low for 30 seconds(!) before boot, bad 6560-VIC. Got a couple replacements from kind people on Reddit, but they both had problems as well. First one had a bad clock divider, so no clock to the CPU. Second one can't seem to consistently drive the address bus to get character data and is a thermal fault. Cold Spray helps, but not for long. Makes me worried that all VIC-I chips' days are numbered like the PLA on the C64 or the CPU on the C16/Plus4.
40:00 people have been working on rudimentary IC fabrication at home but perhaps before that becomes feasible to produce retro replacement parts at scale we might get companies starting to provide cheap custom FlexICs (ICs on flex-cable type substrate, they already made an implementation of 6502 on it). all we'd need is for someone to reverse-engineer any missing chips
well flexIC might not be suitable for making replacement parts for silicon circuits in foreseeable future because it can't switch as fast (the flex6502 runs at 71khz) but maybe some day
15:46 - RADAR RATRACE: a game for the ages... the ages between 5 and 10! :)
Developed, as I recall, by HAL! Not sure if Satoru Iwata worked on it like he did the Galaxian clone-notclone, though.
Great video as always. Funny thing: at 30:47 you typed something like “as* f**k”😂🤣😂
28:12 I remember reading that they made changes to the Commodore 64's KERNAL over the years, and one change was that in some versions the entire color ROM was initialized to light-blue (so if you were to POKE a character into an empty space, it would be visible), but in others it defaulted to the background color and only made the visible characters light blue. I guess you've only ever run this test on machines that used the first style.
I have a merch suggestion - it's making a deoxit that socket shirt (I have a light blue "It Freakin Works" shirt - Spring sucks with their printing process as the ink rubs off easily even with my chewed down not sharp for anything fingernails but Spring is shipping me a new one)
The S logo seems to be the Synertek logo, not Signetics. I don't think Signetics actually did 6502/6522 stuff. The scary thing is the 6522 is still made as the WDC 65C22 and the N variant is directly compatible! (edit) It seems WDC also make a 65C02 which is compatible
8:03 a rare gem indeed!
I have vacuum sealed one of my c64s hoping to perserve it for a few more decades! 😁🤞
The MiSTer project addresses this, as FPGAs are literally programmed using Hardware Description Languages. And someday (perhaps when patents and trademarks expire), the chips will be both be able to be re-created from VHDL or perhaps some other hardware description language, or likely the more practical solution; continue to run as designed in "virtual" FPGA solutions. One additional issue for Commodore computers are the keyboards, but even now, TheC64 can be hacked to work as a USB keyboard.
the 6502 always is a problem on VIC20 (well if you have experience, that is) and not so much on 1541. ROM also is known to go bad on VIC20, 6560/61 also seems to go bad sometimes. Just usualy VIC20 stuff.
I'd like to see the dead chips under an x-ray imager to see if we can see the broken bits visually.
74LS645 is a direct new replacement for the 245s.
I suspect sometime in the next few decades we will have 3d printed ASICs and can recreate these old chips from their photomasks
It could be that the guy who sent you these, populated the sockets with questionable parts.
He did say these were left overs after he repaired his. You could place appropriate TVS on the supply lines, to make sure you don't get over-voltage spikes.
over voltage spikes would be a power supply problem... best thing to do there is use modern power supplies rather than the old junk that they shipped with
Is rammy still with you?
Stick a mirror on your bench. That'll help reaching for around the back of stuff you're working on and is a pro-tip for working on CRT monitors
In what sense are these *_mother_* _boards..._ where do you insert the daughter boards?
Why are you still using a red pen on the chips when you have white ones? It is much easier to see a white mark on a black chip.
I think old chips are a lot more sensitive than newer parts, and perhaps some brands/batches are worse than others.
MOS had questionable quality in general, from their wafer repair technology that turned defective chips marginal. And the mentioned 6526 are especially flimsy by nature.
@@Breakfast_of_Champions yes! I had said to my screen “that’s from their secret sauce that improved yields” :)the older chips had to be perfect to pass inspection so it makes sense they’re hardier.
Vic 20 were does cassette players plug in?
The fact that the song is out of tune in Rat Race makes it even better. You said it annoyed you as a kid but I still kinda laugh at the notion that your parents probably hated it more and regretted buying it. I can imagine one or both sitting down to relax when they hear the first three notes in the distance and then cringing because they know the out of tune rendition of Three Blind Mice will repeat for what feels like forever. Did you use it as a tool/weapon to get more games? Did that cartridge mysteriously disappear at times or suddenly not work for some reason? Or perhaps Rat Race hastened the purchase of a C64. Yes, I can imagine it all.
23:34 funny how the two dead chips on that board both had blue stickers 🤔
Happy Bunny-Day !!! 🐰🐰🐰